URBAN ENERGY ANALYSIS BASED ON 3D CITY MODEL FOR NATIONAL SCALE APPLICATIONS R. Nouvel 1 , M. Zirak 1 , H. Dastageeri 2 , V. Coors 2 and U. Eicker 1 1 HFT Stuttgart – zafh.net, Germany 2 HFT Stuttgart – geo-informatics, Germany ABSTRACT In this paper, we present a methodology based on 3D city modelling to manage a realistic energy analysis of the building stock, building per building, at a very large scale (national application for instance). This methodology is tested on the City of Ludwigsburg and its more than 14.000 buildings. The influences of the data availability and quality on the model accuracy are analysed, for both geometrical and semantical information data. This paper is concluded by exposing some technological trends and policy needs to improve the accuracy and potentials of this methodology. INTRODUCTION Accountable for around 80% of the oil, gas and coal world consumption, urban metropolises are the lead contributors of greenhouse gas production, a main driver of climate change, despite covering only 2% of the Earth’s surface. A rapid transition of urban areas towards energy efficiency and adaption to challenges created by climate change are greatly required. In this context, virtual 3D city models, storing geometrical and semantic data of whole cities, have shown huge potentials in the fields City planning, Environment and Energy, from flood risk simulations to solar potential analyses (Solar Atlas Berlin, 2010). In parallel, the number of cities represented in 3D city models is increasing exponentially, while the investment costs and time required to build these models are decreasing as new automatic data collection technologies such as LiDAR (LIght Detection And Ranging, remote sensing technology measuring distance by illuminating surfaces with a laser) are developed. Some urban energy analysis based on virtual 3D city model have already been realised at local scale for some city districts like in Berlin (Carriòn et al. 2010, Kaden et al. 2013), Karlsruhe and Ludwigsburg (Nouvel et al. 2013). The data quality of these city models are very variable, depending on the available public database (provided generally by the municipality), and the information data collected on- site. Moreover, as they rely mostly on specific (non- standardised) data structure defined locally, they are not applicable to other cities and regions. In this paper, we introduce a methodology of urban heat demand analysis that enables the calculation at national scale of the building heating demands, based on 3D city model (available for all of Germany since 2013), and on two federal databases: theALKIS database (ALKIS, 2014) and European census data. This methodology is first detailed and tested on the whole city of Ludwigsburg (14.000 buildings). Then, the uncertainty of the model is investigated, analysing the influence on the simulated heating demand of building information data. The influence of the CityGML Level of Details, not the same in all cities and regions, is studied as well. In a last section, we describe some technological trends and policy needs which could contribute to the improvement of the accuracy and level of realism of this national-wide energy model. METHODOLOGY The urban energy analysis described and tested in this paper is based on an integrated process using a virtual 3D city model. This integrated process is implemented on the urban simulation platform SimStadt, whose development is ongoing in the Project of the same name (SimStadt, 2014). 3D city model The OGC Standard CityGML (Groeger et al., 2012) has been selected for the modelling of 3D building data. CityGML is an open, multifunctional model that provides a basis for 3D geospatial visualization, analysing, simulation and exploration tools. In recent years, these virtual 3D city models, storing geometric and semantic data of entire cities have shown huge potentials in the fields of city planning, environment and energy, and are increasingly used for these applications. A considerable advantage of CityGML in comparison to other 3D city model formats is its spatio-semantic model, which specifies object modelling in different levels of detail. Due to this, it is an excellent database for heating demand analysis of existing building stocks, since the level of building parameter Fifth German-Austrian IBPSA Conference RWTH Aachen University - 83 -